After the passage of the Patient Protection and Affordable Care Act in March 2010 and the affirmation of its constitutionality by the Supreme Court in 2012, key decisions about the implementation of health care reform are now in the hands of states. But our understanding of these decisions is hampered by simplistic sortings of state directions into two or three simple, rigid categories. This article takes a different approach — it tracks the variations in relative state progress in implementing Medicaid expansion across a continuum of activities and steps in the decision-making process. This new measure reveals wide variation not only among states that have adopted Medicaid expansion but also among those that have rejected it but have also made progress. We use this new measure to spotlight cross-pressured Republican states that have adopted Medicaid expansion or have prepared to move forward and to explore possible explanations for implementation that extend beyond a simple focus on party control.
Editor's Note: Thanks to funding from the Blue Shield of California Foundation and the Robert Wood Johnson Foundation, JHPPL has begun the coordination of an Engaged State Health Reform Research Network to bring together people from different backgrounds (practitioners, stakeholders, and researchers) involved in state-level health reform implementation to inform and extend health reform across the United States. A network website will document implementation projects across the country, workshops will be held, and JHPPL will publish essays under this new section based on findings emerging from network participants. All essays in the section will be published open access.
—Colleen M. Grogan
The battle over the passage and constitutionality of the Patient Protection and Affordable Care Act (PPACA) of 2010 was fought with the soaring language of inspired poets: the implementation of health reform creeps forward — trench by trench — with the fixed bayonets of budget projections, benefit designs, and regulatory fine-tuning. Although the minutiae of implementation are not prosaic, they are the soft underbelly of PPACA's legislative intent.
The PPACA rests on an “access puzzle.” Its original intent was to usher in a historic expansion of access to health care to 32 million Americans. But this bold promise clung to an unstable outcropping of national and state initiatives. Widening insurance coverage for individuals and small businesses was stabilized by new exchanges, which states could choose to establish or could defer to the federal government to operate. Expanding Medicaid to Americans below 138 percent of the federal poverty line was the primary motor for driving greater access; this group accounts for about two-thirds of the new population covered by the PPACA. But the US Supreme Court's June 2012 decision upholding the PPACA's constitutionality and allowing the exchanges to go forward also stripped the teeth from the PPACA's forcing mechanism to aggressively encourage states to expand Medicaid. Specifically, the Court declared the Medicaid expansion optional: Washington could withhold new funding for states that did not proceed with the new programs, but the federal government was barred from touching existing payments. Disparities in state adoption of Medicaid's new programs were inevitable given different political and policy conditions, and the Court's decision only increased their likelihood and magnitude.
Will Medicaid be expanded as intended by the health reform legislation without the threat of federal funding hanging over states? Answering this critical question leads us to ask: what progress are states making in implementing Medicaid reform? For states not already committed to the expansion or those that are proceeding tentatively because of legislative resistance, how well positioned are they to move forward if they decide to reverse course or to accept more attractive options from Washington such as using Medicaid funding to purchase private insurance? Answering these questions requires moving beyond the current crude gauges that divide states into two columns based on whether they have signaled their intent to proceed or not and toward more nuanced measures that place states along a continuum of adoption based on their relative progress in the legislative and administrative processes of implementation. Tracing the process of state adoption makes it possible to study a continuum that not only ranges from stubborn opposition to full implementation but also includes the consequential points in between — states that have not yet agreed to expansion but have taken planning and capacity-building steps as well as states that agreed to adoption but have not yet equipped themselves to implement.
Although the fluidity of state decisions on Medicaid expansion makes it difficult to develop a rigorous explanation that incorporates the multiple factors driving real-time state decisions, it may be helpful to begin raising questions about potential influences on state adoption of Medicaid expansions. The powerful impact of partisan polarization on state decisions about Medicaid expansion in 2012 and 2013 has received attention in the media (Pear and Cooper 2012) and in early scholarly accounts (Oberlander 2011; Jones, Bradley, and Oberlander 2012; Francis and Francis 2010). There is a clear tendency of Democratic-controlled states to embrace reform while their Republican counterparts oppose it. Nonetheless, there is an important and potentially instructive group of states in which Republicans control the levers of power and have decided to adopt Medicaid expansion or to take steps to prepare for the expansion. These states appear to be “cross-pressured” — though political party orientation tilts them against adoption, other dynamics are pushing them to embrace it. What are these other countervailing forces that may be modifying the pressure of partisanship?
We begin by outlining a more robust measure of real-time state decision making on Medicaid expansion in 2012 and early 2013 that adopts a “process approach” as opposed to the “generosity approach” used in past studies of Medicaid development. We then begin to raise questions about the effects of state decisions in 2012 and 2013 on Medicaid expansion that move beyond party control of state governments to account for the cross-pressured Republican states that are making progress.
What Is Happening?
Important research on Medicaid's development since its inception in 1965 has constructed useful measures to track changes over time in the relative generosity of its benefits (Moffit, Ribar, and Wilhelm 1998). This body of research identifies sources of state differences in generosity (Buchanan, Cappelleri, and Ohsfeldt 1991) as well as the interaction of Medicaid benefits with physician services (Cohen and Cunningham 1995) and with the participation of beneficiaries from other programs such as Supplemental Security Income (Yelowitz 1998).
One particularly helpful line of research studies the distinctive features and political dynamics of Medicaid's separate programs. This research shows, in general, that interest group politics drives policy regarding benefits, while traditional welfare politics infused with partisanship guides policy toward eligibility (Grogan 1994, 1999). The generosity and dimensional approaches to Medicaid expansion over its half century bring welcome rigor and sophistication.
Although these approaches offer an important tool kit for future research, they are not particularly well suited for tracking real-time state progress in implementing the PPACA's reforms of Medicaid. For this targeted purpose when the outcome of benefit design, eligibility, and other features are still forming or are yet unknown, we need measures that trace the process of implementation through executive and legislative branch decisions. Rather than asking about the generosity of established state programs, the immediate question is, how far along are states in deciding to adopt new Medicaid provisions?
The Kaiser Family Foundation and other organizations have been dutifully tracking whether states support, oppose, or remain undecided about expanding Medicaid under PPACA. Figure 1 presents Kaiser's trichotomous sorting of states. The general pattern is that about half of the states (twenty-six) are supportive, approximately one-third (seventeen) are refusing to go forward at this point, and seven are weighing their options (Kaiser Family Foundation 2013e).
This trichotomous classification of states is helpful for providing a snapshot of adoption decisions across the country but misses consequential variations in state decision making that obscure institutionally and politically important developments and the potential for future change. To capture variations in real time, we have developed an additive measure of each state's progress in implementing the new Medicaid provisions by tracing three components of the executive and legislative processes through April 2013. (The appendix describes the scale in detail.)
The first relates to gubernatorial statements, budgets, or collective decisions of the legislature and the governor to adopt PPACA's Medicaid expansion (scored as 3 points to signify its weight) or to reject it (scored as -3 points). Twenty-six received a score of 3, eighteen states registered -3 points, and six states tallied 0 points for failing to act.1 Gubernatorial commitments and collective decisions are, of course, primary indicators of state implementation; this is why the Kaiser Family Foundation and the Advisory Board, whose codings we rely on, focus exclusively on gubernatorial or legislative statements or action. Scoring the intent of lawmakers, however, neglects state planning and preparatory steps; our next two components capture these and are omitted by Kaiser's measure and those of others.
The second component of our measure of Medicaid expansion gauges state planning steps and specifically tracks receipt of federal grants. State receipt of federal support indicates a substantial level of seriousness about reform: the state has established an internal process to prepare a plan for reform that the governor submits, and it gains the resources to equip itself to plan and implement the PPACA. For instance, Tennessee remains undecided about Medicaid and therefore is treated like other states that have not declared their intent (such as Kansas) by Kaiser's measure. In reality, Tennessee has received three level-one grants worth more than $8 million to push along its planning, whereas Alaska has not received any federal funding, reflecting its disinterest in reform and lack of preparation. In our measure, states are awarded 1 point for each level-one and level-two grant they received, creating a range from a maximum of 4 points in Nevada to 0 points in fifteen other states. The 1-point metric recognizes state progress while reserving the highest weight (3 points) for states where governors or legislatures signal their intent to adopt Medicaid reforms.
Our third measure traces the concrete changes in Medicaid policy that states took in 2013 — expansion of benefits, simplification of the application and renewal process, and decreased co-payments. These policy changes reveal the tangible disposition of states toward Medicaid reform, increased access, and a user-friendly process. To capture the full range of responses, our scale also tracks state decisions in 2013 that hinder or block progress — benefit reductions, cuts in program eligibility, and reductions in long-term care. Although these tangible steps reveal state dispositions for or against Medicaid reform, none are decisive in the adoption of the new provisions and therefore are accorded 1 point each, positive or negative.2
An additive scale is appropriate because we are measuring the similar underlying phenomenon of state adoption of Medicaid — lawmaker decisions, reform planning, and policy changes. It provides a simple summary of state activity at a point of flux. An alternative approach would be to model each of the three components separately rather than summing them up; this would make it possible to treat each component as a fundamentally different process. Any advantages accorded to this approach would be offset by the inability to treat some components as continuous (lawmaker decisions) and by the loss of parsimony (three variables rather than one). This assessment may well change over time as more states adopt the PPACA's expansion of Medicaid (reducing the importance of assessing relative progress toward implementation) and as distinctive, well-formed dimensions of the new benefits form and become institutionalized.
Figure 2 vividly displays the advantage of our process measure of the relative progress of states moving toward adoption of new Medicaid provisions as compared with Kaiser's simple sorting. Kaiser's classifications in figure 1 treat the twenty-six states in which lawmakers signed off on the expansion as a monolith and similarly lump together the seventeen states opposed to moving forward. But the reality is that each clump of states varies enormously in planning and preparation — New York and Florida are both moving toward adopting Medicaid expansion, yet New York is much farther along. These differences have consequences in terms of each state's commitment to reform and likely future directions.
Of particular importance for questions about the scope of Medicaid expansion, variations among states currently expressing opposition or uncertainty about reform make it clear that there is not a solid bloc rejecting reform — despite similar-sounding rhetoric. Indecisive states like Kentucky and West Virginia as well as ambivalent oppositional states like Virginia are substantially farther along in reform than the stubborn refuseniks Texas, Louisiana, and Wisconsin. Indeed, our measure of Medicaid adoption registers the relative degree of intense opposition in the latter states by assigning them negative scores — a crucial variation that is missed in the invariant classification of states currently resisting the PPACA's new provisions provided by others. Appreciating the range of progress among states not initially signing onto Medicaid expansion may help anticipate decisions by states after January 2014 to accept the new programs.
What explains the variations across states? The fluid nature of Medicaid expansion raises challenges for developing a rigorous explanation for these state variations, but we can begin to frame questions, tentatively consider potential influences, and spotlight an important set of cross-pressured states where Republicans controlling levers of power have made progress toward adoption. What types of pressures may be leading these cross-pressured Republican states to move forward even if incrementally?
Why Do States Vary in Medicaid Adoption?
We take an initial step to explore the variations across states in implementing the PPACA by exploring the relationship of Medicaid adoption with some possible influences on state decision making. We begin by considering the association of Medicaid adoption with political party control — one of the most pervasive influences on US government policy making at the state and federal levels (McCarty, Poole, and Rosenthal 2008; Aldrich and Battista 2002). We proceed to consider the potential association of Medicaid adoption with three other potential (if more moderate) influences — state affluence, previous policy trajectories, and administrative capacity. The purpose of these explorations is to frame questions and jump-start discussion of explanations for state Medicaid adoption; future research will need to sort out direct and indirect influences on state Medicaid decisions and control for the possibility of spuriousness.
Political Party Polarization
Health reform was born in a bubbling cauldron of party vitriol. Not a single Republican in Congress voted in favor of the PPACA's passage, while nearly all Democrats did. That split carried over into a two-year legal battle over its constitutionality as well as the 2010 and 2012 elections. It should be no surprise that the battle lines between Democratic supporters of the PPACA and Republican opponents reappeared in the states and their implementation decisions (Jacobs and Skocpol 2012).
To investigate the partisan nature of Medicaid's development we created an additive measure of party control of state government in which scores increased based on Democrats holding the governorship and one or both legislative chambers. Democratic control of both the legislature and governorship was scored as 6 points, 3 points were awarded when they held the majority in one branch, and 1 point was given when the control of the legislature was split.3 (See the appendix.)
Figure 3 uses a scatter plot to display the distribution of states. Party control is closely correlated with Medicaid expansion (r=.69; p<0.01). States with greater Democratic power are generally moving fastest and farthest in implementing Medicaid, whereas Republican control corresponds with relative inaction or slower progress on the reform. New York, California, Maryland, Washington, and other states with unified Democratic control are reform leaders. Illustrating a bit of partisan flag waving, New York governor Andrew Cuomo heralded the Supreme Court's favorable ruling for ratifying the “leadership of President Obama and his administration … [to] provide access to health care to millions of Americans nationwide and more than one million New Yorkers” (Cuomo 2012). By contrast, the unified Republican governments in Texas, Kansas, Oklahoma, and South Carolina are determinedly opposed to implementation. This partisan patterning of state implementation of the PPACA has been observed by other researchers (Oberlander 2011; Jones, Bradley, and Oberlander 2012; Francis and Francis 2010)
Patterns in figure 3 do suggest, though, that partisan politics may not be sufficient to account for the variation across states. For instance, unified Democratic control is associated with reform progress in California and Maryland but not in West Virginia, which lags behind states where Republicans wield power.
Conversely, states like Arizona, Ohio, Florida, and Michigan exhibit unified Republican government, and their governors are pressing to move forward with Medicaid. Nevada and Arizona offer still another variant of states where Medicaid is progressing even though Republicans could have blocked it. These patterns are telling given the political risk to Republican lawmakers who cross partisan battle lines: influential conservatives in the GOP nomination process have threatened to support primary challengers against lawmakers who moved forward on the PPACA (Erickson 2012).
Why are states with Republicans controlling all or some of the lawmaking branches moving ahead, and why are some states with significant Democratic control lagging? Answering these questions confidently will require additional research, but these questions do raise an important line of future research into what cross-pressures may be moderating the tractor beam of party control in certain states. Indeed, previous research on state health policy making suggests grounds for moving beyond political parties as a singular driver of Medicaid adoption (Grogan and Rigby 2009; Miller and Blanding 2012). We next consider several other factors that may impact Medicaid adoption.
The dire economic straits of certain states offer a particularly compelling account for adopting Medicaid's new programs. The math is, indeed, gripping — deliberately so. Washington covers all the cost of expansion for the first three years and, afterward, continues to pick up an unusually large portion of the administration and benefit costs — 96 percent of costs, on average (Angeles 2010; Kaiser Family Foundation 2010b; Bachrach and Jacobs 2012). In addition to these direct subsidies to states, the enlarged program will pick up costs and segments of the uninsured population that had landed previously on state budgets, thereby reducing state and local governments' uncompensated care expenses.
Federal funding during unsteady economic and budgetary times may be especially attractive to states with particularly strained circumstances; these states need the funds more, and the federal matching formula provides them with more money as compared with better-off states. Previous research finds that less-affluent states are sensitive to fiscal incentives. For example, fiscal incentives tend to prod less-affluent states more than their wealthier counterparts to increase Medicaid enrollment in managed care (Kim and Jennings 2012).
The decisions of some states seem to confirm the impact of fiscal enticements. Arizona's Republican governor, Jan Brewer, devoted parts of her 2013 State of the State address to making a pragmatic case for adopting Medicaid expansion: “Saying ‘no’ to this plan would not save these federal dollars from being spent or direct them to deficit reduction. No, Arizona's tax dollars would simply be passed to another state — generating jobs and providing health care for citizens in California, Colorado, Nevada, New Mexico or any other expansion state” (Office of the Arizona State Governor 2013). Instead, she asked her state to “weigh the evidence and do the math. With the realities facing us, taking advantage of this federal assistance is the strategic way to reduce Medicaid pressure on the State budget. We can prevent health care expenses from eroding core services such as education and public safety, and improve Arizona's ability to compete in the years ahead.” Illustrating the political risk of fiscal realism, she was criticized by the National Review for “exemplifying that unfortunately common strain of Republican leadership that is uncompromising in rhetoric but opportunistic in reality” (National Review Online 2013).
Do relative state economic conditions correlate with state Medicaid progress, as widely assumed? To begin to examine this potential relationship, we created a measure of state economic circumstances by using per capita state income. Specifically, we created a measure of per capita personal income for all states based on population data from the US Census and state-level personal income data from the Bureau of Economic Analysis (2013), averaged across the four quarters of 2010.4
Figure 4 displays state Medicaid adoption and state affluence, and it raises several interesting patterns for future research. First, the modest bivariate relationship (r=.29; p<0.05) contradicts the assumption that finance and specifically state economic need specifically are, as a general rule, driving state decisions regarding Medicaid reform. The intriguing pattern in figure 4 is that states with high per capita income — Connecticut, New York, and Massachusetts — are moving ahead with implementation, while the least well-off states (Mississippi and Idaho) are not. This correlation conflicts with the presumption that economic need would act as a sledgehammer driving states to adopt Medicaid's “good deal.” Indeed, state income may lead to a quite different dynamic: states with the weakest economies may be especially sensitive to even the marginal additional costs required to receive federal funding as well as to questions about the reliability of future funding from a Washington mired in serial fiscal crises.
Of course, generous federal funding is a factor in state decision making (as Arizona's Brewer claims), and lobbyists for hospitals and other stakeholders persistently make this clear to state lawmakers (Ario and Jacobs 2012). While budgetary demands in an era of stringent budgets are a consideration, relative economic circumstances may not, as a general rule, dictate state decisions about Medicaid. Figure 4 raises challenging questions; multivariate analyses will be necessary to control for potential confounding effects.
Where You Stand Depends on Where You Sit
A diverse body of research suggests that the trajectory of established policy informs how government decision makers understand their policy options. Investigations of organizational change and American political development point to prior policy legacies as a factor in national and state decision making (Jacobs and Skocpol, forthcoming). Growing bodies of research from Social Security and GI programs to assistance for the poor and federated social policies show that new government initiatives not only extend new services but also create or recast structures of governance — the “political ground of practices, rules, leaders and ideas” that constitute political authority in a particular place and time (Orren and Skowronek 2004, 20; see also Campbell 2003; Skocpol 1992; Mettler and Soss 2004). In terms of state social policy development, Paul Pierson (1995) argues that systems of federated governance influence social policy development by “chang[ing] the power, preferences, and strategies of social groups [and generating] … new institutional actors” (Pierson 1995: 472). Put simply, policy creates both politics and the context within which lawmakers determine feasible options.
In addition, a number of studies of health policy making reach broadly similar conclusions. For instance, Ae-sook Kim and Edward Jennings (2012) report that state experiments and policy development generate policy learning about Medicaid managed care, which influences the extensiveness of their programs. Henry Glick and Scott Hays (1991: 847) find that the process of creating and then amending policy affects the direction and innovativeness of state decision making.
Previous research suggests, then, that state decision making toward adopting the PPACA's Medicaid provisions may be influenced by prior policies toward low-income people and the uninsured — especially policies toward eligibility and benefits that were established before the PPACA. We created an additive scale of past policy decisions to track program generosity. The scale's focus on program generosity shares the orientation of earlier Medicaid studies (Grogan 1994 and 1999; Moffitt, Ribar, and Wilhelm 1998; Buchanan, Cappelleri, and Ohsfeldt 1991; Cohen and Cunningham 1995; Yelowitz 1998) and is quite distinct from our measure of state adoption decisions, which tracks the legislative and regulatory processes. This scale sums whether and the degree to which the past decisions of states expanded Medicaid to pregnant women, working parents, the medically needy, childless adults, and to more than 90 percent of children (84.8 percent are covered, on average, through SCHIP as of 2009). A handful of states have adopted all these changes; others have not adopted any of these expansions. The precise coding is described in the appendix.
State decisions to move toward adopting the PPACA's Medicaid expansions are moderately well correlated with the generosity of past policy decisions to widen access (Pearson r=.43; p<0.01). Figure 5 suggests, as anticipated by previous research, that differing trajectories in policy generosity correspond with the pace of Medicaid expansion. Longtime leaders in expanding access — Vermont, New York, Connecticut, and others — are also farther along with adopting reform. At the other extreme, states (such as Texas) that stood out for the highest uninsurance rates in the country owing to their unusually restrictive eligibility policies are also most opposed to adopting the Medicaid expansions.
The most intriguing set of states in figure 5 is clustered toward the middle. These states are moving ahead with Medicaid even though Republicans are in control there. Michigan, New Jersey, and other states with a history of pursuing policies to widen access have extended that trajectory by declaring their intent to adopt the new Medicaid expansions. These states are not as far along as Democratic-controlled states, but their recent actions appear to fit into historic patterns of policy activism to expand access. Although additional research is required, this initial correlation raises the possibility that the gravitational pull of policy history (along with other factors) is cross-pressuring states and moderating the effects of political parties on contemporary decisions about health reform.
Previous research on historical institutionalism suggests a compelling reason for this pattern: once established, policies generate both identities and groups that equate their interests with programmatic continuation and expansion and generate resources to mobilize beneficiaries. For instance, voting rates among seniors were comparable with those of other age groups before Social Security was established. Once Social Security was in place, seniors became the most politically engaged age cohort as they identified themselves as a group, connected their interests with the program, and became the target of ambitious organizations (such as the American Association of Retired Persons) seeking to mobilize them and grow their membership (Campbell 2003). The implication for health reform is that the policy trajectory of widening access may have generated identities, interests, and resources that predispose states — in conjunction with other factors — to support adoption of the PPACA's Medicaid expansion.
Institutional Capacities of State Health Policy
What states can do administratively is another factor that may moderate the impact of partisan control on state implementation of Medicaid. According to research by organizational economists and institutionally oriented political scientists, policies that create administrative capacity foster durable and expanding development along the same path (Skocpol 1992; Pierson 2000; Skocpol and Ikenberry 1983). Administrative capacity has both general effects in boosting the confidence of authoritative policy makers and politically powerful allies as well as specific effects in equipping government with the tools to design, adopt, and implement effective programs.
Research on health policy similarly points to the significance of administrative capacity to determine eligibility, process enrollments, assure payments, and monitor quality of care (Holahan et al. 1998). For instance, more effective procedures and resources to enroll Medicaid recipients in managed care in Oregon and Massachusetts produced better results than in states with weaker administrative capacity (Gold, Sparer, and Chu 1996). Administrative effectiveness appears to boost the confidence of government officials and to contribute to additional policy change. While some research develops measures of overall institutional strength (e.g., Barrilleaux, Feiock, and Crew 1992), these studies of Medicaid develop measures that are targeted to specific responsibilities related to the program (Holahan et al. 1998) and particular new initiatives (Gold, Sparer, and Chu 1996).
To explore the potential effect of administrative capacity on Medicaid adoption, we constructed a measure of state resources and procedures related to the PPACA's new provisions. Measuring state administrative capacity in the area of health policy is daunting; past analysis offers no clearly defined measure, with most tailored to narrowly drawn research questions. Our measure is designed to capture both the capacity of states in a common area of responsibility (insurance oversight) as well as more specific capabilities related to aiding the poor and vulnerable such as policing against Medicaid fraud and opening high-risk pools to the medically needy. (See the appendix.) This measure of established organizational capacity reveals wide variation between the weak resources in Alabama and the greater capabilities in New York, California, and Washington. This is a relative measure of the variations across states; most or all states complain (often privately) about lacking adequate capacity to adopt historic reforms in a compressed time period.5
Figure 6 displays several suggestive patterns related to the moderate association of Medicaid adoption and state administrative capacity (Pearson r=.42; p<0.01). States like New York and California that enjoy relatively stronger administrative structures were farther along in putting reform into place, while states sharing Alabama's weak administrative capacity were lagging most in establishing the new reforms.
There is a general overlap of administrative capacity and party control — Democratic states boast greater administrative muscle than Republican states. It is possible that blue states that tend toward greater government activism would, as a result, build more impressive administrative tool kits than traditional red states that eschew government.
The particularly intriguing pattern is that of states characterized by Republican political power and notable administrative capacity — Florida, New Mexico, New Jersey, and Michigan. Additional research will be necessary to explore whether partisanship in these states is being moderated by administrative capacity, contributing to more reform progress than might be predicted by party control alone (Jacobs and Ario 2012).
Although a causal connection has not been established, variations in state capacity to regulate and intervene in medical care and its financing may affect the confidence of government officials and key stakeholders regarding the feasibility of the PPACA's Medicaid expansion.
Conclusion: Tilting toward Reform
We have introduced a new approach to studying Medicaid development that focuses on the real-time decisions of states to adopt or plan for the PPACA's new provisions. Our process approach offers an advance over trichotomous measures that neglect large and consequential variations among states that are officially signaling their intention to adopt the new programs and among those that are refusing to move forward or have remained undecided. Understanding the nature and sources of these variations will have enormous implications for whether Medicaid fulfills the burden placed on it to expand access to more than 20 million Americans.
We also identified a series of intriguing puzzles about sets of states that defy the influence of party — whether Democratic-controlled states like West Virginia that are holding back on Medicaid expansion to this point or states with GOP control that are proceeding. State affluence, past policy trajectories, and administrative capacity may influence state adoption of Medicaid. Future research is needed to sort out causal relationships and identify how states are being pressured.
Examining patterns of state Medicaid adoption raises several broader implications. First, the 2010 PPACA has launched a process that is in its early stages and will evolve substantially as politics shift — giving life to new stakeholders, challenging existing ones, and spurring potentially significant institutional developments that alter current patterns and open new paths. These often slow-moving patterns of creation and destruction in reaction to landmark social policy are not new. The historic occasion of Medicare's passage in March 1965 triggered both celebration by reformers and renewed mobilization by the armies of resistance — doctors prepared to strike and Southern states threatened not to treat African Americans (Jacobs 1993, 2007). In time, well-organized groups of constituents, insurers, providers, and medical suppliers formed to press their interests in ways that substantially expanded Medicare and contributed to enormous changes in the health care system that were not imagined in 1965 or were blocked during the legislative process in the early 1960s (as was the case with the prospective payment system that started to be adopted in the early 1980s). Similarly, the Social Security Act was passed in 1935 with restrictions that excluded about two-thirds of all African American workers (upward of 80 percent in parts of the South) and more than half of all working women, and then it faced several decades of sustained efforts to repeal or quietly smother it. It took decades to make it inclusive, and it was not secure until the 1960s and, most clearly, the early 1970s, when benefits were indexed to inflation. In other words, the story of Medicaid's expansion is just beginning. We should expect change.
Second, reducing state decisions regarding Medicaid expansion entirely to partisanship is understandable given the PPACA's origins, but it may leave us with an incomplete account that diminishes our capacity to understand, most consequentially, the cross-pressured states. The endorsement of reform by Republican governors in Arizona, Florida, Ohio, and elsewhere may reflect the influence of fiscal pragmatism, administrative capacity, and policy trajectories. Our speculation is that as the partisan fury over the PPACA continues to recede, this may open the way for these enduring dynamics to play a greater role in pressuring Republican lawmakers to adopt Medicaid's new programs.
Third, our search for patterns among states seems to suggest practical guidance for states that are undecided or are initially opposed to Medicaid expansion. The prospects for reform proceeding in these states may be enhanced by anchoring them in the foundation of existing policy and by augmenting their technical capacity. Partisanship is a constraint, but it is not a death sentence. Enhanced administrative resources and attention to established policy trajectories may offer strategic levers and points of intervention to moderate the depressive effects of party control where they exist.
Appendix: Coding of State Medicaid Expansion Measure
State Decision to Implement PPACA-Related Medicaid Reform
3 points: Gubernatorial statements, budget announcements, or legislative action indicating that the state plans to implement reform
0 points: State undecided about the implementation of Medicaid reform
-3 points: State has made a clear indication through gubernatorial statements or legislative action that it will not participate in reform
Level-One Grants Awarded
3 points: States that have received three level-one grants from the federal government to implement PPACA reform
2 points: States that have received two level-one grants from the federal government to implement PPACA reform
1 point: States that received one level-one grant from the federal government to implement PPACA reform
0 points: No level-one grant awarded to state
Source: Center for Consumer Information and Insurance Oversight n.d.
Level-Two Grants Awarded
1 point: States that received one level-two grant from the federal government to implement PPACA reform
0 points: No level-two grant awarded to state
Source: Center for Consumer Information and Insurance Oversight n.d.
Medicaid Benefit Expansion
1 point: State legislature's passage of any expansions in Medicaid benefits in 2013 as determined by Kaiser Health Facts, not tied to PPACA benefit expansion provisions
0 points: No state benefit expansions in 2013 as determined by Kaiser Health Facts
Source: Kaiser Family Foundation 2013d
Simplification of Medicaid Application and Renewal Process
1 point: State legislative actions to streamline and simplify Medicaid enrollment procedures making it easier for beneficiaries to obtain and maintain coverage as determined by Kaiser Health Facts for 2013
0 points: No state efforts to simplify enrollment or reenrollment procedures
Source: Kaiser Family Foundation 2013d
Medicaid Long-Term Care Expansions
1 point: Expansion of Medicaid coverage for long-term care during fiscal year 2013, includes expansions related to PPACA and nonmandatory changes in state provisions
0 points: No expansions of long-term care in the state during fiscal year 2013
Source: Kaiser Family Foundation 2013d
Medicaid Benefit Reductions
-1 point: State legislature's passage of reductions in Medicaid benefits for 2013, state response to increasing Medicaid expenses caused by increases in enrollment
0 points: No state reductions in Medicaid benefits for consumers
Source: Kaiser Family Foundation 2013c
State Cuts in Eligibility for Medicaid
-1 point: State legislature's passage of eligibility reductions in any specific benefits of Medicaid for the 2013 fiscal year
0 points: No reductions in eligibility for consumers of Medicaid
Source: Kaiser Family Foundation 2013c
Long-Term Care Cost-Containment Efforts
-1 point: Limits of Medicaid coverage for long-term care during 2013
0 points: No new limits on long-term care in a state
Source: Kaiser Family Foundation 2013c
Coding of State Medicaid Policy Trajectory
Coverage for Low-Income Adults
2 points: Medicaid benefits available for low-income adults as determined by Kaiser Health Facts
1 point: Limited coverage for low-income adults as determined by Kaiser Health Facts
0 points: No coverage for low-income adults as determined by Kaiser Health Facts
Source: Kaiser Family Foundation 2013a
Coverage for Working Parents
2 points: Available to working parents above 99 percent of the federal poverty level as determined by Kaiser Health Facts
1 point: Available to working parents between 50 percent and 99 percent of the federal poverty level as determined by Kaiser Health Facts
0 points: Only available to working parents below 50 percent of the federal poverty level as determined by Kaiser Health Facts
Source: Kaiser Family Foundation 2013a
Coverage for Pregnant Women
2 points: Available to pregnant women above 185 percent of the federal poverty level as determined by Kaiser Health Facts
1 point: Available to pregnant women at 185 percent of the federal poverty level as determined by Kaiser Health Facts
0 points: Available to pregnant women below 185% of the federal poverty level as determined by Kaiser Health Facts
Source: Kaiser Family Foundation 2013b
Medicaid Payments per Enrollee (States Awarded Points by Quartile)
3 points: States with Medicaid payments per enrollee between the 75th and 100th percentiles for fiscal year 2009 (the most recent year available from Kaiser Health Facts)
2 points: States with Medicaid payments per enrollee between the 50th and 75th percentiles for fiscal year 2009 ( the most recent year available from Kaiser Health Facts)
1 point: States with Medicaid payments per enrollee between the 25th and 50th percentiles for fiscal year 2009 (the most recent year available from Kaiser Health Facts)
0 points: States with Medicaid payments per enrollee in the lowest quartile for fiscal year 2008 (the most recent year available from Kaiser Health Facts)
Source: Kaiser Family Foundation 2009b
State Percentage of Eligible Children in SCHIP
3 points: States with 90 percent participation or higher in SCHIP
2 points: States with SCHIP participation between 87.4 percent and 89.9 percent
1 point: States with SCHIP participation between 84.9 percent and 87.3 percent
0 points: States with SCHIP participation at or below 84.8 percent, the national average as of 2009
Party control is an additive scale in which states received points based on the party in power in a state's executive and legislative branches. Three points are given for a Democratic governor, 3 points are given for a Democratic legislature, and 1 point for a split between the two branches. There were 0 points given for a Republican governor or legislature (National Conference of State Legislatures 2012a, 2012b).
Composite administrative capacity is a cumulative measure based on the following: 1 point in the small group market if states guarantee and issue all products, have special rules for groups of one, have state-imposed limits on rating, have health insurance subsidies and up to 1 point for state authority to review rates. The same number of points was available for the same subjects in the individual market. One point was also awarded for the presence of a high-risk pool program, the high-risk pool being open to Health Insurance Portability and Accountability Act–eligible individuals, the high-risk pool being open to medically needy individuals, and the high-risk pool being open to health coverage tax credit–eligible individuals. One point was also awarded for state-mandated coverage of infertility treatment and eating disorder coverage in the individual and small group markets. One point was also awarded for having a false claims act, antikickback laws, self-referral laws, qui tam provisions in the False Claims Act, and the Claims Act meeting Deficit Reduction Act requirements. Half a point was given if the False Claims Act applied only to Medicaid. Three points were awarded for a state having an all-payer claims database, up to 1 point was given (based on Kaiser categories) for federal hazard preparedness funding, and 1 point was given for state conversion coverage in small firms, state restrictions against balance billing in preferred provider organizations and out-of-network providers in health maintenance organizations. The maximum score is 28.5 (Kaiser Family Foundation 2009c, 2009d, 2010a, 2010b, 2010d, 2010e, 2011a, 2011b, 2012a, 2012b, 2012c, 2012d, 2012e, 2012f, 2012g, 2012h; All-Payer Claims Database Council 2013).
State affluence is a measure of state per capita personal income that uses the 2010 Census for population data and data from the Bureau of Economic Analysis for quarterly state personal income. All states data were calculated by dividing the average quarterly personal income in 2010 by the total population in the state for 2010 (Bureau of Economic Analysis n.d.).
1. Although our measure partly relies on Kaiser's coding, its classification differs in several respects. Specifically, Kaiser lists South Dakota and Wyoming as weighing their options, whereas we code them in opposition. Kaiser codes Virginia in opposition, and we code it as weighing its options because both legislative chambers amended the state budget to allow for Medicaid expansion discussions even though the governor has not proposed it. Wyoming was coded as not participating because Governor Matt Mead recommended that the state not participate in the program. Finally, we coded South Dakota as not participating because Governor Dennis Daugaard announced in his state budget address that the state does not plan on participating in the program (Advisory Board Company 2013).
2. The scores for the three components of our additive measure reflect their relative importance. The authoritative nature of lawmaker decisions distinguished them and was accorded 3 points compared to 1 point for discrete policy actions by states and for planning grants. Alternative metrics, for example, increasing the weight of grants or implementation, produced scales with similar overall variations, though the specific ranking of individual states varied slightly. This scale was chosen for analysis because of the symmetry in points awarded between decisions for and against reform implementation.
3. Virginia scored 1 point because each party holds the same number of seats in the state senate.
4. We used the average quarterly per capita income because it allowed us to account for any variations in a given quarter in a state. Second-quarter income produced similar results.
5. Interviews with state health reformers were conducted by Jacobs.